LONDON ELECTORAL HISTORY STEPS TOWARDS DEMOCRACY...
Transcript of LONDON ELECTORAL HISTORY STEPS TOWARDS DEMOCRACY...
LONDON ELECTORAL HISTORY –
STEPS TOWARDS DEMOCRACY
2.1 POLL BOOKS
Prior to 1832, the qualifications for being a metropolitan elector varied
between the different constituencies. Hence historians need to examine
with care the nature of the franchise (see section 3) but, even before that,
the nature of the sources. Indeed, it can be said that source-criticism is the
essence of scholarship.1 This section accordingly analyses the sources used
for creating the LED. It not only explains how poll books came to exist
(and to be preserved), but it also assesses their reliability and how they can
be used by historians.
Historians working on the period prior to the rise of public opinion
polling at the end of World War II must infer the political behaviour of
individuals from aggregated data.2 Whilst this is far less costly than the
collection of individual-level data, there are a number of problems inherent
in it. Ballot papers in British parliamentary elections are aggregated within
each constituency prior to counting, making the constituency the smallest
possible area for spatial or ‘ecological’ analysis.3 No social data are
included on the ballot papers, and constituencies are rarely coterminous
with those census registration districts for which social data are available.4
Meanwhile, the overall ‘ecological’ analysis of political data (as exem-
plifying a local or regional community) suffers from the considerable
problem that relationships at the aggregate level often do not hold true at
the individual level.5 For example, national-supremacist candidates may
thrive in constituencies with many immigrant electors, not because the
immigrant voters support them, but because their arrival stimulates the old-
established population to express resentments arising from competition for
housing and employment.6
2.1.1 Poll book sources generated by open voting
By contrast, historians working on the period of open voting have in poll
books a rich source of individual-level behavioural data. In the English
2 LONDON ELECTORAL HISTORY
context, poll books are the records of voting in parliamentary elections
prior to the introduction of the ballot in 1872.7 Although printed poll
books are a peculiarly English phenomenon,8 open voting was not
confined to the British Isles. It was customary in parts of the United
States until 1896.9
And it persisted in Denmark until 1900; in Prussia
until 1918; and in Hungary, strikingly, until 1938.10
Poll books contain the names of the voters and details of the candi-
dates for whom each polled. The simplest poll book is thus a list of the
names of those who voted for a particular candidate.11
Many also contain
supplementary data, most commonly lists of those who voted for other
candidates.12
Further information recorded in county poll books may
include where the voter lived, the location of his freehold, and the name
of his tenant. Meanwhile borough poll books commonly record the
voter’s address and occupation. This was the standard format before
1832. All voters in this period can be accurately described by male
pronouns (as already noted in website section 1.9.1). That state of affairs
was sustained by strongly entrenched custom rather than by legal
proscription, so that women were de facto excluded from the suffrage,
even if they wielded influence behind the scenes.
Occasionally, the poll books recorded also the voter’s religious
affiliation;13
his family relationships;14
his tax assessment;15
the name of
his livery company;16
or how he had voted at a previous election.17
Sometimes, too, the books included informal lists of qualified electors
who did not poll.18
Moreover, after 1832, there are frequent examples of
borough poll books that indicated the voter’s qualification for voting,
noting whether he was an ‘old’ or a ‘new’ elector, as well as providing
information about the property that qualified him to vote.
Poll books have survived in both manuscript and printed form.
Manuscript poll books may be either the poll clerks’ copy,19
or
contemporary check books maintained by the candidates’ agents and
inspectors.20
The role of the poll clerk was a responsible one, usually
taken very seriously.21
Meanwhile printed poll books were published for
a variety of reasons. In preparation for a scrutiny, they provided an
opportunity for neighbours or fellow liverymen to check for fraudulent
voting,22
and they acted as pro-forma canvassing books for a subsequent
election.23
Furthermore, some were distributed after the election by
candidates to their supporters;24
and some may have been purchased by
voters wishing to see their own names in print.
The use of the poll books for electioneering purposes indicates that
2.1 POLL BOOKS 3
there was often a degree of organisation attached to the campaign
process, and that groups or parties had a life-span beyond the moment of
just one election. However, their waxing and waning roles should not be
taken to imply that there were two long-term and permanently opposed
national political parties, with rival ideologies. This had been assumed,
for example, by G.M. Trevelyan, who wrote of ‘the two perennial groups
labelled Whig and Tory … [which] continued as the strongest and most
lasting element in our public life from the days of Clarendon and
Shaftesbury to the days of Salisbury and Gladstone.’25
Rather the
rhetoric of partisanship was seen as divisive, and was avoided by most
candidates throughout the period. Many candidates consciously tried to
remain neutral. An example was Admiral Hood who wrote in 1783,
when considering his prospective political career, that: ‘I shall ever most
carefully and studiously stand clear, as far as I am able, of all suspicion
of being a party man’.26
Indeed, as already noted, no party labels were
ever used at the polls throughout this period. In response to the
complexities of allegiances and those who tried to stand above the
partisan fray, the LED does not pre-allocate party labels to candidates (as
already noted in section 1.7 Descriptive terminology), thus allowing
users to interpret the evidence for themselves.
Historians have generally worked with the printed poll books.27
There
are a number of reasons for this. First, the destruction in 1907 of the
manuscript poll books formerly preserved among the public records at
the Crown Office has left many elections recorded only in printed
editions.28
Secondly, printed poll books are legible, complete, and
accessible. Yet not all contested parliamentary elections led to printed
editions of poll books. Often poll books for elections in counties and
large cities were not printed,29
presumably because of the costs
involved.30
The introduction of electoral registration after the Reform Act
of 1832 led to the first arrival of a hybrid document, the marked electoral
register. In such examples, the printed electoral registers were annotated
by hand with indications of the candidates for whom each voter polled.
For the creation of the LED, both printed and manuscript poll books
have been used, together with marked registers where these have
survived.31
All three of these kinds of document record electoral
behaviour, and must be distinguished from canvassing books, which
show the declared voting intentions of the electors. Although a few
canvassing books survive, these have not been incorporated into the
LED.32
Simple lists of voters survive for two elections: these are rare but
4 LONDON ELECTORAL HISTORY
have similarities with poll books, except that they only record who voted
and contain no information on their political choices.33
Nonetheless all
information about turnout (whether an elector voted or not) is of great
importance, so these lists together with lists of unpolled London
liverymen have been incorporated into the LED.
2.1.2 The advent and survival of poll books
The origins of the practice of recording votes are obscure. Early elections
were decided by a show of hands; indeed, the show of hands long
continued to determine the outcome of many elections. Then, in the early
seventeenth century, some poll lists were compiled, either as an official
record of voting or as evidence to support a petition to the House of
Commons in controverted elections.34
By the 1660s manuscript poll books were becoming commonplace.35
In 1696, fresh legislation36
required that in county elections some formal
record be kept, in all cases of contests that could not be determined by a
simple show of hands:
In case the said election be not determined upon the view ... but
that a poll shall be required for the determination thereof, then the
said sheriff ... shall forthwith there proceed to take the said poll ... and
... shall appoint such number of clerks as shall to him seem meet and
convenient for taking thereof; which clerks shall all take the said poll
... and to set down the names of each freeholder, and the place of his
freehold, and for whom he shall poll. And be it further enacted, that
every sheriff, under-sheriff, mayor, bailiff, and other officer, to whom
the execution of any writ or precept shall belong for the electing
members to serve in parliament, shall forthwith deliver to such person
or persons, as shall desire the same, a copy of the poll taken at such
election, paying only a reasonable charge for writing the same.
Initially, it was only at county elections that there was direct authority
for returning officers to appoint poll clerks. But in 1725 London gained
the statutory right to have poll clerks appointed, when it was declared that
‘The presiding officer … shall, in case a poll be demanded, appoint a
convenient number of clerks to take the same’.37
By another statute of 1746,38
which applied to cities that constituted
counties of themselves (and were thus outside the jurisdiction of the
2.1 POLL BOOKS 5
surrounding county), the sheriff was to allow ‘a check book for every poll
book of each candidate’, implying that the sheriff had power to appoint
clerks to keep poll books. But no provision was made for the permanent
preservation of poll books, although these records continued to provide
crucial evidence in the event of a disputed election.39
Indeed, not until the
statute of 1843 was the preservation of borough poll books officially
required.40
By the time that borough poll books were required to be deposited, the
run of printed poll books for London had already ended, whilst Westmin-
ster had already acquired a fine series of electoral records among the
muniments of the Westminster Sessions of the Peace. Manuscript poll
books from London’s parliamentary elections are rare: that of the by-
election of 1781 is the only one so far encountered. Westminster’s early
eighteenth-century poll books had already disappeared by 1789.41
But,
with the exception of the election of 1807, the series for Westminster is
complete for the period between 1774 and 1820, although a number of
poll books have subsequently decayed.42
Quite possibly the poll books
were handed down from one high bailiff to his successor before finding
their last resting-place.
England’s earliest printed poll book (from the Essex by-election of
1694) pre-dates the statutory requirement for the taking of polls, and
many pre-date the requirement for the preservation of poll books. But the
taking of polls rapidly stimulated the appearance of printed editions: of
which about 2,200 survive.43
Historians have largely relied upon these
printed editions, and a number of printed editions survive for the
metropolitan region. Manuscript poll books survive for many Middlesex
and Westminster elections for which no printed edition was published, as
well as for London’s parliamentary by-election of 1781. Hence, in
compiling the LED, these have been used, together with surviving polls
from non-parliamentary elections, to provide additional material when
appropriate. Yet the completeness of printed poll books gives them
advantages over the fragmentary survivals of manuscript poll books,
despite potential errors in transcription and typesetting, so the core of the
LED rests upon political choices that were made publicly and later
published.
It may further be noted that the study of voting behaviour from all
poll books has the great advantage of studying evidence that was
recorded for the purpose for which it can be analysed: for votes were
cast in order to be counted.
6 LONDON ELECTORAL HISTORY
2.1.3 Poll books and the study of electoral behaviour
Early attempts at the social analysis of poll books pre-date the
application of computers to the discipline. In a trail-blazing article of
1960, Rudé examined the electoral behaviour of individual voters over
successive elections to reveal ‘a close concordance between those voting
Proctor-Cooke and Luttrell on the one hand, and among those voting
Wilkes and Glynn [the radicals] on the other’.44
It was Rudé’s great good
fortune to discover in the Middlesex poll books of 1768-9 not only a
source that allowed him to explore the social context of the Wilkites but
also one in which the electoral data were already linked by the hand of
an unknown contemporary scribe.45
Rudé’s spatial analysis of voting
showed that rural Middlesex was less radical than the urban parishes.
Hence Wilkite support was concentrated in ‘the populous urban parishes
lying to the east and north of the City’.46
Furthermore, Rudé’s equally
pioneering use of rate books for social analysis demonstrated that a
majority of the more substantial voters polled for Proctor or Luttrell,
while the vote for Wilkes, the anti-establishment hero, was drawn from
the lesser freeholders.47
Meanwhile, Victorian county poll books from the period after the
Reform Act were adopted as a key source by D.C. Moore, to demon-
strate the continuance of what he termed as ‘deference communities’ that
followed aristocratic leadership.48
Analysis of nineteenth-century poll
book data was continued by J.R. Vincent, whose prodigious efforts to
make sense of dozens of electoral contests led him to reject a dichoto-
mous class conflict between Liberals and Tories but to identify instead a
myriad of social and occupational interests in play, depending upon the
local circumstances.49
Other studies have also explored local and
regional variants, noting the two-way interactions between political
leadership and voter concerns.50
Before the first application of the computer to the discipline, historians
had shown great ingenuity in using poll books for a variety of purposes,
such as J.H. Plumb’s calculations of the size of the electorate under the
Stuart kings.51
However, the range of questions that could be examined
systematically was immeasurably augmented once the power of the
computer was applied. Here the pioneering work came from W.A. Speck
who first formulated, in relation to the county electorates of the early
eighteenth century, the questions of behaviour, loyalty, and turnover that
continue to dominate the subject.52
2.1 POLL BOOKS 7
So was advanced the methodological sophistication of a field that has
come to be known as historical ‘psephology’, appropriating a new term
launched into circulation by D.E. Butler in 1952.53
Thus, at much the same
time as Speck, Michael Drake first used a computer to perform record
linkage on poll book data. This technique enabled him to construct
‘psephological trees’ for the electors of Ashford, Kent, in the mid-
nineteenth century.54
Similar work of considerable methodological
sophistication by Mitchell and Cornford led to a rudimentary ‘flow of the
vote’ model for the nineteenth-century borough of Cambridge.55
This form
of analysis was characterised by a shift from a single poll book towards the
linkage of two or more poll books,56
using increasingly sophisticated
algorithms. Other have also linked poll book data to other sources. Thus
R.J. Morris combined polling data with information from directories and
voluntary associations to explore the dynamics of middle class formation in
newly enfranchised Leeds.57
Meanwhile, J.A. Phillips made significant methodological and substan-
tive contributions. His Electoral behaviour in unreformed England showed
a rise in partisanship in the later years of the eighteenth century.58
Later
studies by Phillips revealed an increasingly participatory electorate,
especially in relation to partisan alignments at the time of the Reform Act
of 1832.59
His approach was matched by O’Gorman’s finely nuanced
synthesis of much recent work on the local context of electoral politics,
together with an analysis of a wide variety of borough constituencies
throughout the Hanoverian period.60
O’Gorman argued that the unreformed
electoral system continued to function satisfactorily until the eve of the first
Reform Act, and that in many constituencies it satisfied the aspirations and
needs of the parties to electoral relationships.
Contested elections could assume particular significance in local
communities, when the chosen candidates of rival elite groups solicited
electoral support from electors who were their social ‘inferiors’.61
There
was a potential moment of social inversion, when power was potentially
taken from the rulers and wielded by the ruled. At the same time, it was
also argued that elections ultimately shored up social harmony, by making
the electors feel themselves to be courted and part of the system. It was a
point made in 1785 by Archdeacon Paley – who officiated in Carlisle, a
classic pocket borough in the hands of a local patron:
Popular elections procure to the common people courtesy from their
superiors. That contemptuous and overbearing insolence, with which
8 LONDON ELECTORAL HISTORY
the lower orders of the community are wont to be treated by the
higher, is greatly mitigated where the people have something to give.
The assiduity with which their favour is sought on these occasions,
serves to generate settled habits of condescension and respect; and …
whatever contributes to procure mildness and civility of manners,
towards those who are most liable to suffer from a contrary behav-
iour, corrects the pride … the evil of inequality, and deserves to be
accounted amongst the most generous institutions of social life.62
Yet rituals of rebellion and inversion may do more than simply
reinforcing existing social structures. By offering an alternative model of
society, they may themselves be part of a process of change.63
The
pioneer in studying the social meanings of eighteenth-century campaign
rituals is undoubtedly O’Gorman;64
and there is scope for similar work in
application to nineteenth-century elections. On the theme of ‘inversion’,
it may be noted that to this day the occasional humbling of a Cabinet
minister on election night can remind otherwise sceptical electors that
their votes do count for something. Divisions are thereby revealed; and
also settled by the result of a vote. In the eighteenth century, there were
traditional rituals after the votes were declared, whereby the winning
candidates were publicly acclaimed (for example, by being chaired
around the streets). Such measures marked the attempt at communal
reaggregation after a liminal period of electoral conflict.65
Despite its growing analytical yield, the study of historical psephology
has not been without its critics. For example, it has been argued that the
smaller contested boroughs for which polls were printed were not
necessarily representative of the large constituencies in which the majority
of the pre-reform electorate lived.66
Hence historians may risk studying
atypical electors. And the motivations of those who voted remain unknown.
Thus, while historians who analyse poll books tend to see most voters as
acting purposively and exercising political or social choice, other historians
have retained the older view that depicts an essentially venal electorate,
trying to sell its votes to the highest bidder.
Furthermore, it is justly noted that in this period MPs formed relatively
loose and flexible party groupings in the Commons, and that they did not
contest parliamentary elections under national party labels.67
Governments
rose and fell from power in the years from 1700 to 1850 without direct
reference to the outcome of a general election. Moreover, it cannot be
assumed that the results of local electoral contests, with their selective local
2.1 POLL BOOKS 9
franchises, can be projected onto public opinion as a whole. Poll books
were not opinion polls; nor were they intended to be.
Nonetheless, the constitutional framework, with all its oddities and
historical complexity, was the guiding template for the political system,
both before and after 1832. The results of elections were taken seriously, as
a manifestation of the links between the centre and the localities, and as a
declaration, especially via the votes case in the large ‘open’ constituencies,
of the public ‘mood’. It therefore remains appropriate to study how
politicians and electors acted, within that framework. Even the most ardent
reformers wished to improve, rather than to abolish, the representative
system, by which MPs represented a local constituency in the ‘high
parliament’ of the kingdom. For historians, then, poll books contain
behavioural data which stem from many people’s conscious decisions, and
can be studied as such.
That individual voters’ decisions were made within complex local and
personal circumstances remains true – as it does of voting behaviour today.
The act of open voting in the years 1700 to 1850 indicated that each voter
at least gave a publicly witnessed declaration of support for one or more
candidates. But people’s secret motivations were, of course, never stated.
The case of John Moody, the proprietor of a shoe warehouse in Westmin-
ster’s Carnaby Street, offers a pertinent reminder that things were not
always as they seemed. Moody was a member of the London Correspond-
ing Society and secretary of his friend Horne Tooke’s Westminster
campaign in 1796. And Moody voted consistently against the government
candidates. In Westminster in 1802, he plumped for Fox, and as a
Middlesex freeholder in the county contest in the same year, he plumped
for the radical Burdett. These votes marked him as a determined reformer,
who declined to cast his second vote for a moderate candidate. Yet it is now
known that Moody was a government spy who had infiltrated the
metropolitan radical movement. As ‘Notary’ he reported to the government,
his true identity known only to his ‘handler’ Richard Ford and a few other
people.68
Moody’s radical voting profile was thus consistent with
concealing his role as a spy but was misleading about his secret allegiance.
Yet even then the historian has no way of knowing why Moody agreed to
inform upon his colleagues. He may have acted out of sincere belief, or for
money, or under pressure of some sort.69
Records of voting behaviour thus do not open ‘windows into men’s
souls’. In the era of open voting, however, the poll books did record what
people did publicly; and also what was the outcome of many such public
10 LONDON ELECTORAL HISTORY
actions. This evidence can be used, therefore, to survey individual
behaviour in its public aspect as well as in its aggregate outcome, which
constituted the sum of all individual choices. Voting required electors to
crystallise their opinions, no matter how complex those might be in private.
Nothing can be done to alter the overall pattern of data survival,
although historians can still hunt for hitherto undiscovered sources. The
LED includes a mixture of previously known and previously unknown
election results for the metropolis in the years 1700-1850, now collected
together for the first time. It is presented for historians to explore the
challenge and problems of party classification, as has been already
undertaken for the Westminster election campaigns of Charles James
Fox.70
And it further invites historians to consider the formidable but
interesting difficulties of translating occupational labels into social classes
and economic groupings.71
And, above all, it is open for any or all other
forms of geographical, social, political, and psephological analysis.
By tipping the balance back towards the large urban constituencies, the
LED thus recognises that a live constitutional tradition was being enacted,
from a variety of voter motives, at the geographical heart of the country’s
political system. The unreformed electoral system was far from perfect. But
what was sought by the reformers in unrepresented Manchester before
1832, as well as by all campaigners in other unenfranchised boroughs, was
the opportunity to participate in a manner similar to the electors of the
metropolis.
2.1.4 Poll books used in the LED
Much of the LED is derived from printed poll books, while some comes
from contemporary transcripts. But the LED is not a simple transcription
of the extant documents. Rather, it is an edited text, a representation of
historical events that were transmitted orally at the hustings and then
recorded by two textual processes: the typesetting of the printed poll
books and the later entry of data into the database. That stage has been
accompanied by a process of textual emendation to supply a best
reading, especially in the Street and Occupation fields, of this
refractory material. Few of the poll books from which the LED was
created are the poll clerks’ Ur-texts that preserved the sequence in which
electors polled, while even those created from the original poll clerks’
books have additional problems of missing data.
Many extant poll books used to create the LED have survived in a
2.1 POLL BOOKS 11
single source only. In these cases there is no question of determining a
preferred source. Nonetheless, even a list derived from a single source
may be defective in some respect. To take one example, the copy of
Anon., Lists of the liveries of London (1701) in BL has been severely
cropped in binding, with the loss of some records. For the LED,
therefore, its evidence has been taken from a less mutilated copy in the
Bodleian.
Other poll books exist in variant editions and recensions, and many of
the lists used are problematic. Manuscript poll books may suffer from
missing data, so it has sometimes been necessary to create a patchwork
text from different sources. For example, the Westminster poll book of
1784 used to create the WHD lacked about a third of the records for St
James’s parish. For the LED, a contemporary transcript of the St James’s
poll book72
was used in preference, with the occasional gaps in this
record being supplemented by the WHD edition, taken from the poll
clerks’ books.
As in the WHD, other records remain incomplete. The printed West-
minster poll book of 1780 lacks occupational data, so for the parish of St
Anne and for the united parishes of St Margaret and St John it was decided
to use information for this election that was found in the original poll
clerks’ books.73
Meanwhile, the loss of the original poll book from St
George in the Westminster by-election of August 1788 was remedied by
data from a contemporary transcript, doubtless once the property of an
election agent.74
And missing data from the Westminster election of 1796
have been supplemented from two contemporary transcripts. That for the
united parishes of St Margaret and St John is apparently complete.75
But
those for St Anne, St Martin, and St Paul with St Martin-le-Grand are
themselves incomplete, listing only plumpers for Fox, plumpers for Horne
Tooke, and votes split between Fox and Horne Tooke.76
Early poll lists were sometimes issued containing details of those
voters who had polled for one slate of candidates. These lists, which were
generally issued by the other slate of candidates as an invitation to check
the bona fides of the voters, present considerable problems with split
votes. The poll of the London parliamentary election of 1713 was issued
as two such lists, which have been edited together in the LED to form a
single table. Similarly, the poll of the London parliamentary election of
1722 first appeared as two discrete lists.77
This election, contested by two
slates of three candidates, is fortunately also to be found as the single poll
book that was the preferred source for the LED.78
12 LONDON ELECTORAL HISTORY
Alone among the elections represented in the LED, the poll for
London in 1713 is represented by a composite table, pieced together
from two surviving lists of voters.79
Each of these lists contains the
names of those who gave votes to one or more of its slate of candidates.
Moreover, each discrete list is preserved, the Whigs having 4,024 names
in Poll ID 27 and the Tories 4,166 in Poll ID 28.80
The first stage of the
process by which Table PL1713 was created consisted in identifying
and isolating those 3,315 Whigs and 3,478 Tories who gave four votes to
a slate of four candidates. In the second stage, voters whose records
matched between the two lists on the fields of Company, Surname, and
Fname, and who gave not more than four votes split between the two
lists, were identified, and their votes concatenated to create 404 new
records. The third stage consisted of the identification of what appeared
to be matching pairs from the remainder; the two lists were then edited
to enable Stage 2 matching where appropriate to create a further 206 new
records. Ultimately, then, a fourth stage identified those 175 voters who
appeared in one list only, but who cast fewer than four votes.
Meanwhile some printed poll books exist in different or variant
editions. At the simplest level, those additions and corrections in later
issues have been incorporated into the poll book tables in the LED. The
list of those who polled for Sir John Williams in the election for sheriff
in March 1724 exists in two versions; it was not possible to determine
which (if either) list was a revised version. So the list issued in the Daily
Journal, which was a public resource, was used.81
In another case, two
printed lists of those who polled for Charles Goodfellow at the London
parliamentary by-election of December 1724 have survived. Correction
of the spelling of names of some of the liverymen, while leaving
unchanged their position in the alphabetical sort, shows the list preserved
in the National Archives to have been corrected by a contemporary.82
That source has therefore been preferred for the LED, in lieu of the
uncorrected list issued immediately after the election.83
A similar decision has been made with reference to the poll book for
the London parliamentary election of 1768. It was quickly reissued with
additions and corrections, which have been duly incorporated into the
LED table.84
Again, two lists survive of those who polled for Frederick
Bull at the London parliamentary by-election in December 1773; and the
corrected source has been used for the LED.85
And, after the London
parliamentary election of 1784, two lists were recorded of those who
polled for Richard Atkinson. One of these was no longer was unavailable
2.1 POLL BOOKS 13
to historians at the time of data entry,
86 so the other has been incorpo-
rated, together with its companion list of those who polled for John
Sawbridge.87
In other cases, the printed sources have survived in multiple editions,
requiring selection of the apparently most complete examples. Of the
three Middlesex polls published in 1705, the LED incorporates Anon.,
An exact list (1705). And of the three editions that appeared after the
Middlesex election of 1802, the LED has used Anon, Copy of the poll
(London, E. Rider, 1803). In another case, there are rival sources in print
and manuscript: for the Middlesex elections of 1768-9, the LED has
preferred the manuscript poll book transcript at LMA, which contains
contemporary notes linking the entries,88
rather than the printed version
at BL.
Whether manuscript or printed, poll books were inevitably compro-
mised by human fallibility. The imprecise orthography of the poll clerks
and compositors led to renderings in the Westminster poll books such as
‘Cartwright Street’ for Carteret Street, ‘German Street’ for Jermyn Street,
‘Palm Alley’ for Pall Mall. Others included ‘Putney Street’ for Pulteney
Street, ‘Thrift Street’ and ‘Fish Street’ for Frith Street, and ‘Tuttle Street’
for Tothill Street. These obvious inconsistencies have been corrected in the
LED, and other place names have been standardised, though no doubt more
emendations could still be undertaken.
Further standardisation was required because the names of some
streets have changed. Street names were standardised only from the
middle of the nineteenth century; one of the first being the ‘New Road’
which ran from Edgware Road to the Angel in Islington. This stretch of
road had 55 subsidiary names, consolidated in 1857 into ‘Marylebone
Road’, ‘Euston Road’, and ‘Pentonville Road’. For example, in the
Marylebone poll books of 1837-41 there appear the constituent parts of
Marylebone Road: New Road, Harley Place, York Buildings, Salisbury
Place, Cumberland Place, Queen Charlotte Row, Homer Place,
Winchester Row, Albany Terrace, Ulster Place, Allsop Terrace,
Gloucester Place, Lisson Grove South, Middlesex Place and Southamp-
ton Row.89
All of these have been standardised in the LED.
One invaluable source for this exercise is the 1901 List of the streets,
which gives both old and new names within the administrative county of
London. Incidentally, the creation of an index to this work would be a
valuable addition to London’s topographical history.
Sometimes different variants were used concurrently. A modest
14 LONDON ELECTORAL HISTORY
terrace of houses, ‘built to let’,
90 might lie within a known street but be
identified also by the name of the promoter, as in ‘Smith’s Terrace’,
‘Smith’s Buildings’, or ‘Smith’s Rents’.91
Similar names upon roads and
housing are reminders that metropolitan property offered major
opportunities for investors, at a time when investment opportunities were
comparatively constricted. Thus it is no surprise to find that profits from
London’s notorious sex trades also found their way into London
property.92
Variants in nomenclature were also found application to large
thoroughfares as well as to small ones. What is today known as Oxford
Street was identified also as Oxford Road and Acton Road. On other
occasions, a thoroughfare had both a formal and an informal name. Thus
Crown Court, located immediately north of Pall Mall in St James,
Westminster, was known informally as Paved Alley (perhaps to
distinguish it from another Crown Court in the same parish). Meanwhile,
it seems likely that Pissing Alley in the City of London was also known
by a more formal name.93
In other cases, one name superseded another. Hog Lane in St Anne,
Westminster, was superseded by Crown Street and is now known as
Charing Cross Road. King Square in the same parish became Soho
Square. Leicester Fields became Leicester Square. And Torment’s Hill in
St Margaret, Westminster, became St Ermine’s Hill. Occasionally,
something was lost in the nineteenth-century standardisation of street
names: the junction of Brewer Street and Walker’s Court in St James,
Westminster, now at the epicentre of the Soho sex trade, was known for
much of the eighteenth century as ‘Knave’s Acre’.
Nor were all street names unique. In 1856, when the first attempt
was made to impose some order on street naming and numbering across
the metropolis, it was found (as if in a symbolic trial of strength between
the royal houses of Stuart and Hanover) that London contained 48
different streets called ‘Charles Street’ and 46 called ‘George Street’.
Meanwhile, in a further genuflection to monarchy, 38 streets were called
‘King Street’ and 36 ‘Queen Street’. Westminster’s parish of St James
itself contained two King Streets, one (now Kingly Street) near Golden
Square, the other near St James’s Square.
Other names in need of editorial standardisation can be found in the
Middlesex electoral records. ‘Armingsworth’ was replaced with
Harmondsworth, ‘Chinkford’ with Chingford, ‘Chissell Street’ with
Chiswell Street, ‘Endfield’ with Enfield, ‘Grazing Lane’ with Grays Inn
2.1 POLL BOOKS 15
Lane, and ‘Stebon Heath’ with Stepney. Further emendations included
Isleworth for ‘Thistleworth’, Hoxton for ‘Hogsdon’ and ‘Hoston’,
Ickenham for ‘Hickingham’, Laleham for ‘Lalam’ and ‘Laylamb’,
Marylebone for ‘Marrowbone’, Muswell Hill for ‘Muzzle Hill’,
Shoreditch for ‘Show Ditch’, and Twickenham for ‘Twittenham’.
The London poll books presented fewer orthographic problems,
since all but the poll for the by-election of October 1781 came from
printed sources. Nonetheless, eighteenth-century typesetters’ orthogra-
phy offered some amusement: ‘Rude Lane’ has its charm, and ‘Labour in
Rain Court’ sounds even more depressing than the correct ‘Labour in
Vain Court’. The Six Clerks’ Office sounded like a numbered property in
a terrace, as did both Three Crowns Court and Three Daggers Court. No
name, however, caused the typesetters as many problems as did
Marylebone’s small street known as Beaufoy Terrace, which appeared in
numerous variants.
Meanwhile, the old way of identifying a property by a sign long
coexisted with new attempts at sequential numbering of properties in a
street,94
as testified by innumerable advertisements in the London press
throughout the eighteenth century. The Westminster parish rate books,
organised as a cadastral survey (being an accurate street-by-street
listing), began to include street numbers only in the late eighteenth
century; but it was well into the nineteenth century before all parts of the
metropolis followed suit.95
Moreover, even where street numbering
existed, it might be inconsistent: for example, Craven Street, off the
Strand in Westminster, was consecutively numbered along each side.96
Given the overlap between old signs and new numbers, it is not
surprising that the street numbers given in some of the poll books and
rate books seem to bear little relationship to those given, for example, in
Richard Horwood’s Plan of the Cities of London and Westminster
(1807). The process of standardisation followed only slowly.
Thus it should be stressed that the place and street name strings in
the LED are codes that bring together discrete items of data that, in the
historian’s judgment, appear to relate to the same entity. Thus ‘Oxford
St’ in the LED is a code representing the strings ‘Oxford Street’, ‘Oxford
Road’ and ‘Acton Road’ in the original poll books. The principles of
coding, and the record linkage processes thereby facilitated, are
discussed in further detail in section 4.
There were equally bizarre spellings of personal names, but these and
other poll book fields required less editorial intervention than did those
16 LONDON ELECTORAL HISTORY
containing addresses. Surname strings were entered as a single continuous
string, thus ‘Delisle’ or ‘Ocallaghan’. But obvious eighteenth-century
contractions such as ‘Jno
’, ‘Cha’, Josh’, and ‘Geo’ were silently expanded.
Clear typographical errors, such as the ‘Geroge’ who appeared in the
WHD, were corrected.
Occupation strings presented fewer editorial problems, in part
because of the use of the Dictionary table to implement occupational
classification. This allowed such variant spellings as ‘plaisterer’ and
‘plasterer’ to be given the same occupational code. A very few
occupational strings warranted editorial emendation: an improbable
‘haymaker’ in the Shop Tax records became a hat maker, whilst the man
described in the same source with an esoteric combination as ‘chain
maker and carpenter’ was thought to have practised his carpentry skills
in chair making.
The process of standardising, classifying, and interpreting data that
reaches the historian in non-standard forms is a fundamental first step
towards systematic analysis. Comparable processes are undertaken by
historians of all subjects who standardise dates and spelling, and who
group items of information together, by geographical or any other system
of classification.
So the data in the LED amount to an edited version of the sources
upon which it is based, just as those sources are closely related to, but
distinct from, the electoral acts to which they refer. As the publisher of the
Westminster poll book of 1780 noted in his preface:
From the noise and confusion of the hustings during the time of
taking down the names of the voters, and the expedition in printing a
numerous and long-contested poll, like the present, several mistakes
must unavoidably have happened, notwithstanding every possible care
has been made use of to prevent them.97
But it is the poll records, warts and all, which provide historians with
evidence; and with the opportunity to reflect upon the multiple ways that
evidence can be used.
2.1.5 The metropolitan electorates
Voters in eighteenth-century metropolitan elections constituted an elite.
It may have been a broad elite, of liverymen in London, of freeholders in
2.1 POLL BOOKS 17
Middlesex, and of householders in Westminster and Southwark, but it
was nonetheless still an elite. It was not representative of the population
at large. Historians may, however, estimate the extent to which those
who polled were representative of their populations of eligible voters
(liverymen, freeholders, or householders) by examining what would now
be called the turnout. Turnout (whether an elector voted or not)
constitutes the atom of political behaviour; analysis of the political
choices of those who voted constitutes a second-order explanation.
Political commentators frequently (if misguidedly) neglect the impact of
differential turnout, and invoke only second-order explanations of
change in political representation. Issues of ‘who actually voted’
logically come before ‘who voted for whom’.98
Turnout may be estimated for London, for which a population of
eligible liverymen may be found. The practice of making livery lists
goes back to the beginning of the poll book period; indeed, the earliest
list in the LED is a livery list, rather than a poll book. The livery lists of
1701, in which an asterisk was used against the names of each liveryman
who polled, suggest a turnout of over 70 per cent of eligible liverymen.
Meanwhile, the London poll books of 1710 and 1727 record unpolled
electors (some of whom were overseas) and indicate a crude turnout of
almost four-fifths of the eligible electorate. A livery list issued in 1733
recorded not only the names of the members of each of the livery
companies but also the date of their admission.99
Further livery lists were
made in the middle of the eighteenth century: one, dated from internal
evidence to 1750, lists 7,470 liverymen and gives the occupations of
some of them.100
Another, issued in 1751, enumerated 6,535 liverymen,
of whom nearly 4,700 lived within the London, while a further 1,400
lived in Southwark, Westminster, Middlesex, Surrey, Essex and Kent.101
Two livery lists issued in the 1790s enumerated between 9,000 and 9,500
liverymen.102
The practice of issuing livery lists continued until the
beginning of the nineteenth century, when one was published in parts
corresponding to the place of residence of each liveryman.103
However, livery lists are not without their problems. Listings derived
from submissions made by livery companies sometimes neglected
altogether the members of small companies, which may have made no
response to a request for information (note the absence, in Table 7.6.1
below, of any returns for the Tinplate Workers in 1701 and 1750). On the
other hand, livery lists derived from poll books are ipso facto of little value
for estimating turnout. Later eighteenth-century livery lists appear to have
18 LONDON ELECTORAL HISTORY
derived in part from poll books.
104 Nonetheless, it seems likely that
turnouts in the range of three-quarters to four-fifths of the City of London
electorate were common,105
while the infirm and those living at a distance
were the least likely to vote.
Several livery companies maintained lists of their members, which
may be used (as they were in electoral scrutinies) to check the qualifica-
tions of voters. The City Elections Act (1725)106
provided for the return of
lists of liverymen by the clerks of the various livery companies, but it is
clear that this was the codification of an extant practice in respect of
disputed elections. Throughout the eighteenth century the Society of
Apothecaries seems to have printed lists of members entitled to practise
their calling.107
Later the practice of issuing livery lists became more
common, and in the last thirty years of the eighteenth century both the
Drapers and the joint livery of the Tylers and Bricklayers appear to have
issued annual lists, followed in the early nineteenth century by the
Goldsmiths.108
Other companies issued printed livery lists from time to
time, and there are some further survivals of manuscript lists in the
Guildhall library. One possible exercise would be to check London’s poll
books against the livery companies’ apprenticeship admissions registers109
in order to explore both turnout and the age-structure of those liverymen
electors who had undertaken a formal apprenticeship, although the work
involved would be considerable.
Contemporaries were aware of the uncertainties generated by the
absence of any formal listings of potential voters. An attempt was made in
1788 to introduce what would have amounted to a new system of electoral
registration for county electorates.110
But the statute proved to be too costly
to implement, and it was repealed in 1789 without ever having been fully
used.111
There was neither the political nor the administrative will to find a
solution. Thus earlier bills to introduce electoral registration in 1785 and
1786 had failed, as did a later attempt in 1828.112
In the absence of electoral registration, appearance in the Land Tax
registers constituted a de facto entitlement to vote in county elections, such
as those in Middlesex, until the end of the eighteenth century.113
The
financial exigencies of the Napoleonic wars, however, forced Pitt to raise
revenue by Land Tax redemption, allowing people to exchange their tax
liabilities in return for an immediate cash payment. Fresh legislation then
followed in 1802 to preserve the right of voting of the redemptioners,
whose names had been removed from the Land Tax registers.114
As a result
of such manoeuvres, the registers became an increasingly unsatisfactory
2.1 POLL BOOKS 19
record of eligibility to vote in county elections from the beginning of the
nineteenth century.
Meanwhile, in the case of Westminster, there was no list of electors
eligible to vote, before the introduction of electoral registration in 1832. Yet
there is good evidence that, in hotly contested elections, a high proportion
of those who wished to exercise their franchise did so. The Westminster
election of 1784 famously lasted for 40 days, but by the last week so few
came to poll that the constituency may be regarded as having ‘polled out’
with 12,300 voters having polled.
Eventually, electoral registration in 1832 was intended to resolve the
uncertainties; but the new system generated special problems in places like
Westminster, where the traditional electors, in Westminster’s case qualified
as adult male householders ‘paying scot and bearing lot’ (local taxes), were
allowed to retain their right to vote, alongside those newly eligible as
householders occupying property with a rack rent valuation over £10. In
the transition period, some voters were registered twice.115
The LED
contains an example: Samuel Addington of St Martin’s Lane (Westminster)
was registered both as a £10 householder and as a ‘scot and lot’ house-
holder (LED records 6712891; 6714100). In detail, a number of qualifications about the sources also need to
be noted briefly. The problem of missing data is the most intractable.
Another potential problem is the question of known systematic error.
The poll book data for Marylebone within the LED were taken from a
marked electoral register of 1841. The records for 1837 and 1838 thus
suffer from a bias towards those electors (two-thirds of the total
electorate) who survived to be included in the later electoral register. The
remaining third of electors in 1837 and 1838, who had moved away from
Marylebone or died, were not listed and are lost to historians.
Other printed polls also suffer from data loss to a lesser extent. The
published polls for the London parliamentary election of 1784 are for
Sawbridge and Atkinson only. Almost all liverymen polled for one or
other of these, a handful perhaps for both, and the systematic error of
data loss is slight. Again, the printed polls for the three elections for lord
mayor in 1831 do not record a few dozen votes given to minor
candidates. However, as the overwhelming majority of these votes are
likely to have been split with a vote given to a mainstream candidate, the
systematic error caused by their omission is negligible.
Another area where there is a possibility of error, unknown to histori-
ans, is in people’s declarations about their occupation. Certainly, the
20 LONDON ELECTORAL HISTORY
census authorities later in 1891 drew attention to ‘the foolish but very
common desire of people to magnify the importance of their occupational
condition’.116
Such behaviour may have occurred in London a century
earlier when an individual identified as ‘labourer’ can sometimes be found
being translated into a ‘gentleman’ in successive elections – always
assuming that the individual in question was one and the same. There are
also some uncertainties as to the use of terminology, such as the
‘yeomen’ listed in Westminster in 1784, when the clerks appear to have
used the term as something like a synonym for ‘labourer’.117
Yet tests for consistency, when electors are traced from one election
to the next, indicate that many voters gave themselves publicly the same
occupation (or a closely related occupation) over a span of years,
suggesting that there was some verisimilitude to their occupational
declarations – which, under the system of open voting – were made
publicly in the hearing of all. Of course, illegal occupations were not
stated. Those rarely appear in the official censuses either, although the
1881 census did include a ‘retired opium smuggler’, a ‘supposed brothel
keeper’ and a ‘pick a pocket’.118
The occupational representations in the
LED are therefore the respectable face of the metropolitan economy –
and they are, in their own right, a valuable source for historians to
analyse.119
2.1.6 Electoral scrutiny
Fiercely contested elections, in the days of open polling, could be
rumbustious affairs. Crowds, wearing favours in their political colours,
would gather to witness the proceedings, and to shout encouragement for
their own side. It created a form of pressure as well as an atmosphere of
excitement. Some voters, furthermore, were encouraged in their
allegiance by ‘treating’, in which known taverns provided free drinks for
one side or another, at the cost of the candidate. Such excitements did
not happen all the time, because many elections were low-key affairs.
But certainly the proceedings were not conducted with the silence and
decorum of twentyfirst-century elections.
On the other hand, bribery in the form of direct payment for votes
was not common in the large electorates, because the cost would be too
great and the outcome uncertain. The notorious ‘Eatanswill’ electorates,
where votes were traded for money or gifts, were rightly criticised by
reformers before 1832, but such cases applied chiefly to the smallest and
2.1 POLL BOOKS 21
most venal boroughs, not to the large open constituencies.
Indeed, evidence that the London elections in this period were
carried out with tolerable care comes from the investigations, also
known as ‘trials’, that were undertaken by the House of Commons from
time to time, at the request of the defeated candidates. At the scrutiny
before the declaration, some votes would be rejected – not on grounds of
bribery, but on the grounds that the voter was not properly qualified
under the local franchise. At the scrutiny of the votes polled in the
election for a sheriff of London and Middlesex in March 1724, the total
was reduced by about 14 per cent. And the scrutiny that followed the
Westminster election of 1749 reduced the total by about 15 per cent.120
But such proportions were unusually high, following elections that were
unusually fiercely contested.
A scrutiny after the London parliamentary election of February 1701
reduced the total number of votes cast by 440, suggesting that about 110
voters were successfully challenged. Another scrutiny following
London’s parliamentary election in 1727 reduced the total number of
votes cast by 820, indicating that 205 voters, about three per cent of the
total, were rejected. In 1770, the scrutiny following the poll for Lord
Mayor was abandoned with no change in the number of votes.121
At the
scrutiny following the poll for MP for the City of London in 1784 the
total number of votes cast was reduced by less than one per cent, with no
effect on the outcome of the election.122
And, most notably, the scrutiny
undertaken after the disputed Westminster election of 1784 got no
further than considering votes from the relatively small parish of St
Anne, of which only a handful were rejected, before it, too, was
abandoned.123
These results were the more impressive in that there were
repeated accusations of partisanship on the Commons committees that
sat on the petitions.124
One reason for the low level of voters rejected at the scrutinies was
the practice by the candidates of employing inspectors at the polls, who
were able to challenge prospective voters with suspicious credentials.125
In Westminster, for example, these inspectors were numerous and well-
rewarded: in 1774 the agents for Percy and Pelham Clinton employed 15
inspectors who received £20 each for their services, and another who
received £25.126
In the particularly intense campaigns of the 1780s, the
polling of ‘bad’ or dubious votes may have been a matter of policy. George
III, who was implacably opposed to Fox, went so far as to write to Pitt on
13 April 1784 that ‘though the advance made by Mr Fox this day can only
22 LONDON ELECTORAL HISTORY
have been made by bad votes, yet similar measures must be adopted rather
than let him get returned for Westminster’.127
Unfortunately for the agitated
monarch, his chief minister was not able to oblige on this occasion.
Indeed, the evidence of rejected tenders to vote offers some comfort to
historians that by and large the inspectors did their job with tolerable
efficiency. In defence of the process, Charles James Fox noted in the debate
on the Westminster scrutiny that individual voters were checked against the
parish books, and, in dubious cases, enquiries were made at the putative
voter’s place of residence to verify or disallow his claim.128
There may also have been some cases, not of fraudulent voters, but of
personation, whereby one individual illegally voted in the name of a valid
voter. (This practice is remembered in the mythic twentieth-century
exhortation that Ulstermen should ‘vote early, and vote often’). Some
personation must have occurred during this period, but any cases which
escaped the attention of the partisan inspectors at the hustings must
necessarily escape the scrutiny of the historian. At very least, however, it
can be accepted that the names and occupations of those who did vote were
those of qualified voters, whose appearance at the poll seemed in good
order.
Defeated candidates regularly accused their successful rivals of polling
unqualified voters: that was part of the stock-in-trade of eighteenth-century
electioneering. Indeed, the identification of unqualified or otherwise
fraudulent voting was clearly one motivation behind the publication of poll
books.129
Among the Westminster elections such allegations were most
frequent in 1788, when Hood’s petition to the Commons was a comprehen-
sive indictment of electioneering techniques.130
It alleged:
That the returning officer admitted a great number of persons to vote for
the said Lord John Townshend, who, by the ancient usage and custom of
the said city and Liberty, had no right to vote at the said election, and
others who were disqualified from, and incapable by law of, voting at
the said election; and that the names of many persons were received
upon the poll as voting for Lord John Townshend, who, in fact, did not
vote for either of the candidates, but were votes given by other persons,
falsely assuming their names and characters, and several persons were
admitted to poll more than once at the said election; and that, as well
before the said election, as during the time of taking the said poll, many
persons, by bribery, gifts, promises of reward, and other undue and
illegal practices, did corrupt a great number of the voters to poll at the
2.1 POLL BOOKS 23
said election on the part of the said Lord John Townshend.
In the same year, Sir William Young reported a string of allegations to
the Marquis of Buckingham:131
The question is not of title to vote in most cases, but of identity; most
families being at this season out of town, a rascal was found to
personate every absentee. The suborners of perjury not regularly
conferring, very many instances occurred of an absentee being
represented by four or five, all admitted to vote on their mere attesta-
tion.
Yet it seems most unlikely that rational rascals knowingly imperson-
ated someone who had already voted. Party agents routinely kept lists of
those who had polled to check against their canvassers’ lists of promises.
Furthermore, instances of identical poll book records are very rare, and
generally the result of polling by a father and his son having the same
name, address, and trade. William Fox, a victualler of George Street in St
Margaret and St John, apparently polled three times in the Westminster
election of 1788: but this is a very rare instance – and two of these at least
could have been family members with the same name.
In 1788, there were particularly picturesque allegations about the
opposition Whig Townshend’s canvassers in Westminster. They were
said to have identified empty houses, those occupied by women, and the
houses of those who said they would abstain. Strangers were then brought
into Westminster from Hoxton and Shoreditch to assume the identities of
the occupiers of these houses. These impostors were taken to houses in
Covent Garden, where they were dressed and tutored, before being taken to
poll. Others were alleged to have assembled and been victualled at
Sheridan’s Theatre Royal in Drury Lane before creating mayhem at the
hustings. For this, they were said to have received 5s a day, or £3 if they
voted. But it seems unlikely that there were ‘very many instances ... of
women having voted dressed as men’, as Hood’s lawyer alleged.132
The
rhetorical claims, though entirely unsubstantiated, indicate something of
the electoral tactics of creating a buzz of partisanship for one cause –
and allegations of malpractice against the rival cause. Ultimately,
however, those who convinced the inspectors and polling clerks of their
eligibility to vote must also be accepted as being of equal standing by
historians.
24 LONDON ELECTORAL HISTORY
Notes
1 A.E. Housman, ‘Preface to Manilius’ in idem, Collected poems and selected prose,
ed. C. Ricks (Harmondsworth, 1989) remains a classic exposition of the editorial
role, and of waspish wit.
2 Analysis of aggregate electoral data may be traced back to A. Siegfried, Tableau
politique de la France de l’Ouest sous la troisième république (Paris, 1913). For the
nineteenth-century English electorate, see T.J. Nossiter, ‘Voting behaviour, 1832-
72’, Political Studies, 18 (1970), pp. 380-9, and idem, ‘Aspects of electoral behav-
iour in English constituencies, 1832-68’ in E. Allardt and S. Rokkan, Mass politics:
studies in political sociology (1970), pp. 160-89. For the increased sophistication of
multiple regression analysis, see J.A. Turner, British politics and the Great War:
coalition and conflict, 1915-18 (New Haven, Ct, 1992).
3 Voting details may be available at ward level in local government elections. Much of
this information is buried in the pages of local newspapers, and its collection and
publication is a valuable addition to historical knowledge of local politics and party
formation. See S. Jordan, K. Ramsey and M. Woollard, Political representation and
Bristol’s elections, 1700-1997 (Bristol, 1997); and, for London, A. Willis and J.
Woollard, Twentieth-century local election results (Plymouth, 2000), 3 vols.
4 Constituencies were grouped into yet larger units to correspond with census registration
districts in K.D. Wald, Crosses on the ballot: patterns of British voter alignment since
1885 (Princeton, NJ, 1983).
5 W.S. Robinson, ‘Ecological correlation and the behaviour of individuals’, American
Sociological Review, 15 (1950), pp. 351-57. J. Shover and J. Kushma, ‘Retrieval of
individual data from aggregate units of analysis’, in J. Silbey et al., The history of
American electoral behaviour (Princeton, NJ, 1978), pp. 327-39, and G. King, A
solution to the ecological inference problem: reconstructing individual behaviour
from aggregate data (Princeton, NJ, 1997) are more recent contributions to this
thorny problem. But the methodology is abstruse, and it remains doubtful whether
the behaviour of individuals as opposed to groups can indeed be inferred from
aggregate ‘ecological’ data.
6 For a historical example of the ‘ecological fallacy’, in this case the spurious
association between Roman Catholicism and Conservative voting, see Wald, Crosses
on the ballot, pp. 150-1.
7 35 & 36 Victoria, c. 33 (1872). The practice of open voting was famously defended as
the exercise of a trust by John Stuart Mill in his Considerations on representative
government (1861) against the demands of George Grote and others for the ballot.
Since the ballot is necessarily secret (see OED, s.v. ‘Ballot’, to give a secret vote), the
2.1 POLL BOOKS 25
familiar expression ‘secret ballot’ is a pleonasm, although the phrase is so entrenched
that it is hard to avoid. An example of a semi-secret voting system was that used
from 1621-1939 in the Papal Conclaves to elect a new Pope: the cardinal electors
signed their voting papers, albeit folding the document to reveal the vote whilst
concealing the signature. In the later twentieth century, this system was replaced by a
ballot, whereby each elector writes one name on a card which is then folded to confer
secrecy. See F.J. Baumgartner, Behind locked doors: a history of the papal elections
(New York, NY, 2003), pp. 145-6, 236.
8 For locations of poll books, see J.M. Sims (ed.), A handlist of British parliamentary
poll books (Leicester, 1984); J. Gibson and C. Rogers (eds), Poll books: c. 1696-1872:
a directory of holdings in Great Britain (4th edn, Bury, 2008); and E.M. Green, ‘New
discoveries of poll books’, Parliamentary History, 24 (2005), pp. 332-67. R. Chapman,
‘Poll books for the county of Middlesex’, West Middlesex Family History Society
Journal, 11/3 (1993), pp. 16-23, is derivative and adds little.
9 See P. Bourke and D. DeBats, ‘Identifiable voting in nineteenth-century America:
toward a comparison of Britain and the United States before the secret ballot’,
Perspectives in American History, 11 (1978), pp. 259-88; P. Bourke and D. DeBats,
Washington county: politics and community in antebellum America (Baltimore, Md,
1995); and A.G. Bogue, ‘The quest for numeracy: data and methods in American
political history’, Journal of Interdisciplinary History, 21 (1990), pp. 89-116.
10 See J. Elklit, ‘Open voting’, in R. Rose (ed.), International encyclopaedia of
elections (2000), pp. 191-3; and J. Elklit, ‘Nominal record linkage and the study of
non-secret voting: a Danish case’, Journal of Interdisciplinary History, 15 (1985), pp.
419-43. For an overview of this ‘New Political History’, see J.A. Phillips (ed.),
Computing parliamentary history: George III to Victoria (Edinburgh, 1994).
11 Examples of this kind of poll book are remarkably uncommon. One for the
metropolitan region, albeit not for a parliamentary election, is List of pollers for
Richard Hoar, 6 January 1738/9, Guildhall broadside 9/7. The names of the twelve
hundred and forty seven gentlemen of the livery of London … who poll’d for John
Wilkes, esq., at the general election in March, 1768 (R. Withy, 1768) is a piece of
Wilkite ephemera printed in gold on blue waxed paper. A ‘List of the freeholders
who voted for Colonel Luttrell’, from the Middlesex by-election of 1769, is printed
in Oxford Magazine, 3 (1769), pp. 91-5. Another, from Southwark, Surrey, is to be
found in [Hector Campbell], The names of those patriots, in the parish of St John,
that supported Mr Tierney ([London], 1796).
12 See, for example, A list of the persons who polled for Mr Dumello and Mr Tyson at
the late election of common-council men for Bridge ward, London [London, 1740].
26 LONDON ELECTORAL HISTORY
13 Systematic information about religious affiliation is rare, with the significant
exception of poll books from Ireland. Poll books from Abingdon, Oxfordshire, in
1734 and 1754 identify dissenters. While some Quakers may be identified from their
affirmation in lieu of swearing the oaths, it seems likely that returning officers
allowed Quakers through on the nod, rather than disrupting the electoral process
unduly. The right of Quakers to affirm was granted by 7 & 8 William III, c. 34
(1696); it was made permanent by 1 George I, c. 6 (1714); and extended to Mora-
vians (followers of the German Protestant sect) by 22 George II, c. 30 (1749).
14 Indications of family relationships appear to be confined to those who share a common
forename and surname combination. In the LED they are distinguished in a discrete
field as ‘sen’ [senior] and ‘jun’ [junior]. Elsewhere, for example in Newcastle-under-
Lyme, the convention was to distinguish ‘A.B. son of C.B.’ from ‘A.B. son of D.B.’
15 Tax assessments are very rare in poll books, but that from the Bath, Somerset, by-
election of 1855 records the rate assessment of each voter. Meanwhile the poll book
for Newry, Down, in 1868 records the voter’s qualification, the rateable value of his
property, and his religious affiliation. The only metropolitan poll with tax assess-
ments so far encountered is that for the election of alderman for Vintry Ward in
1827, CLRO Misc. MSS. 199.9 – now to be found under LMA classification.
16 All London parliamentary poll books record the livery company of each voter
entitled to poll under the old franchise.
17 Information about previous voting behaviour for entire electorates is rare. Exceptions
include Anon, A state of the polls for the two elections for representatives in
parliament for the county of Cambridge (Cambridge, F. Hodson [1802]), which
includes the voting behaviour of individuals at both the by-election and the general
election of that year. Meanwhile Anon., Acland’s comparative poll book (Hull,
James Acland, 1835), published after the election in June 1835, also included
information on the voting behaviour (classified by party) of those who had also
polled in the Kingston upon Hull constituency in 1832 and January 1835.
18 Information about unpolled electors is necessarily problematic before the introduction
of electoral registration in 1832. Whilst feasible in freeman boroughs, it was impossible
in inhabitant boroughs or scot and lot constituencies like Westminster. But the London
poll books of 1710 and 1727, where the livery franchise was akin to a freeman
franchise, contain useful lists of unpolled electors. Lists of unpolled electors are more
frequently encountered in post-registration borough poll books, for example, that from
Whitehaven, Cumberland, in 1832.
19 Much of the poll data for Westminster between 1784 and 1820 is taken from the poll
clerks’ books.
2.1 POLL BOOKS 27
20 Poll book data for St Anne, St Martin, and St Paul with St Martin le Grand at the
Westminster election of 1796 are taken from a contemporary transcript. This list of
plumpers (single voters) for Fox, plumpers for Horne Tooke, and split votes between
Fox and Horne Tooke, comes from the library of Sir Alan Gardner. This suggests
that candidates also took a close interest in the voting of their opponents even when
there was no possibility of a scrutiny.
21 See contextual discussion in M. Knights, ‘John Locke and post-revolutionary
politics: electoral reform and the franchise’, Past and Present, 213 (2011), pp. 41-86.
22 The printed poll books for the closely fought election for the City of London
chamberlain in 1734, and also those from the contest for lord mayor in 1772, were
published by the committees of the parties opposed to those candidates whose votes
were recorded.
23 See P. Jupp (ed.), British and Irish elections, 1784-1831 (Newton Abbot, 1973), pp. 127-
8, for the use of the printed poll book from the Westminster election of 1818 by
Hobhouse’s canvassers the following year; and WAC E/3349/4/37 for the similar use of
the poll book by Lamb’s canvassers.
24 In nineteenth-century Colchester, Essex, poll books were issued in paper wrappers,
showing the party’s colours, for distribution with the candidates’ thanks to those who
had polled for them.
25 G.M. Trevelyan, The two-party system in English political history (Oxford, 1926), p. 6.
The approach of Namier, especially in the chapter ‘Why men went into parliament’, in
L.B. Namier, The structure of politics at the accession of George III (1929), i, pp. 1-76,
acted as such a solvent upon ideas of permanent party affiliations that some of the most
enthusiastic Namierites tried to eliminate ‘party’ even from periods of great party strife,
causing a later reaction against Namierism.
26 D. Hannay (ed.), Letters written by Sir Samuel Hood (Viscount Hood) in 1781-3 (Navy
Record Society, 1895), pp. 155-6.
27 But there were some exceptions. J.A. Phillips, Electoral behaviour in unreformed
England: plumpers, splitters and straights (Princeton, N.J., 1982), p. 328, reported
supplementing a lacuna among the printed poll books for Maidstone with a manuscript.
The manuscript poll book of the Yorkshire election of 1734 was analysed by P. Adman,
W.A. Speck and B. White in an appendix to J.F. Quinn, ‘Yorkshiremen go to the polls:
county contests in the early eighteenth century’, Northern History, 21 (1985), pp.
137-74. Manuscript poll books were also used in D. Hirst and S. Bowler, ‘Voting in
Hertford, 1679-1721’, History and Computing, 1 (1989), pp. 14-18, and in R. Hall,
28 LONDON ELECTORAL HISTORY
‘Whig party fortunes in the Yorkshire county election of 1708’, Northern History, 32
(1996), pp. 111-32.
28 Although regrettable, this purge was not the disaster that it might appear to have
been. The preservation of borough poll books was only required from 1843, and the
polls of a considerable proportion of contested borough elections were published. For
printed poll books, see Sims (ed.), Handlist of British parliamentary poll books,
together with Green, ‘New discoveries of poll books’.
29 No printed poll book survives for any metropolitan constituency after 1841.
30 Printed poll books in large constituencies were expensive. The Westminster poll books
of 1749 and 1774 were priced at 2s 6d, whilst a publisher’s advertisement for the 1818
poll book offered the work in eight parts at 1s each. See Anon., ‘A collection of
addresses, pamphlets, posters, squibs, etc. relating to the Westminster election, 1818’ (a
BL nonce volume of electoral ephemera). Poll books were a minority taste, and print
runs were probably small.
31 Marked registers have been used for the Marylebone elections of 1837, 1838, and
1841, for the London election of 1848, and for the Westminster election of 1851.
The last printed poll book for London was published after the election of 1837; the
last for Westminster after the election of 1841.
32 A canvassing book for the Middlesex election of 1747 is at Warwickshire RO, CR
132 A253. Meanwhile, a canvassing book for the Paddington parish of Marylebone
in 1832 survives at WAC, T/V/65.
33 A list distinguishing polled and unpolled electors in each livery company survives
for London election of 1701: Anon., The lists of the liveries (1701). A list of voters
also survives for the Middlesex election of 1727 (Alnwick Castle Muniments Room,
YV2b).
34 D. Hirst, The representative of the people? Voters and voting in England under the
early Stuarts (Cambridge, 1975), pp. 122-9. For example, a poll list from Clitheroe,
Lancashire, in 1640 is reproduced in W.S. Weeks, Clitheroe in the seventeenth
century (Clitheroe [1927]), ii, pp. 229-30.
35 M. Kishlansky, Parliamentary selection: social and political choice in early modern
England (Cambridge, 1986), p. 186. M. Knights, Politics and opinion in crisis, 1678-
1681 (Cambridge, 1994) used manuscript poll books from the period.
36 7 & 8 William III, c. 25 (1696)..
2.1 POLL BOOKS 29
37 11 George I, c.18, s. 1 (1725).
38 19 George II, c. 28, s. 6 (1746).
39 Two examples show the process in action. Following a petition arising from the
controverted Westminster election of 1708, the high bailiff John Huggins was
required to lay before the Commons all the poll books, together with lists of the
would-be voters rejected at the scrutiny: CJ, 16, p. 24. Similarly, the Commons in
April 1769 required the Middlesex sheriff to produce the poll books in response to a
petition against the return of Henry Lawes Luttrell: CJ, 32, pp. 385-7.
40 6 & 7 Victoria, c. 18 (1843).
41 TNA PRO 30/8/237, fo 791. See also [Trial of the Westminster election by a
committee of the House of Commons, 4 April – 9 May, 1789] in Lincoln’s Inn
Library, LP 297. This printed Commons sessional paper is not included in S.
Lambert’s reprint edition.
42 LMA Westminster poll books, WR/PP.
43 Sims, British parliamentary poll books, remains the most complete and reliable
guide to printed editions. Some further printed editions of metropolitan parliamentary
poll books, not recorded by Sims, have been discovered in the course of this
research, and are noted at relevant points.
44 G. Rudé, ‘The Middlesex electors of 1768-9’, English Historical Review, 75 (1960),
p. 605.
45 The manuscript poll book at LMA, used by Rudé in this study, contains a
contemporary record that indicated against the name of each voter his choice at each
of the three parliamentary elections of 1768-9. This resource may well have been
compiled for canvassing purposes.
46 Rudé, ‘Middlesex election’, p. 609.
47 Ibid., p. 615.
48 See D.C. Moore, ‘The other face of reform’, Victorian Studies, 5 (1961), pp. 7-34,
idem, The politics of deference (Hassocks, 1976); and P. Joyce, Work, society and
politics: the culture of the factory in later Victorian England (Brighton, 1980). But
see the important qualifications in D. Eastwood, ‘Contesting the politics of defer-
ence’, in J. Lawrence and M. Taylor (eds), Party, state and society: electoral
behaviour in Britain since 1820 (Aldershot, 1997), pp. 27-49.
30 LONDON ELECTORAL HISTORY
49 J.R. Vincent, Poll books: how Victorians voted (Cambridge, 1967), pp. 17-18: for
example, the drink trade acted as a Tory lobby, not universally but particularly in
constituencies where the Liberal candidate was ‘dry’ and pro-temperance.
50 T.J. Nossiter, Influence, opinion and political idioms in reformed England: case
studies from the north-east, 1832-74 (Brighton, 1975), pp. 119-200.
51 J.H. Plumb, ‘The growth of the electorate in England, 1600-1715’, Past and Present,
45 (1969), pp. 90-116.
52 W.A. Speck and W.A. Gray, ‘Computer analysis of poll books: an initial report’,
Bulletin of the Institute of Historical Research, 43 (1970), pp. 105-112; and W.A.
Speck, W.A. Gray and R. Hopkinson, ‘Computer analysis of poll books: a further
report’, Bulletin of the Institute of Historical Research, 48 (1975), pp. 64-90.
53 D.E. Butler, The British General Election of 1951 (1952), p. 1, credits his colleague
R.B. McCallum for coining the term, based upon the Greek word for the pebble
dropped into an urn to vote.
54 M. Drake, ‘The mid-Victorian voter’, Journal of Interdisciplinary History, 1 (1971),
pp. 473-90.
55 J.C. Mitchell and J. Cornford, ‘The political demography of Cambridge, 1832-68’,
Albion, 9 (1977), pp. 242-72.
56 M. Drake, Introduction to historical psephology (Open University: Milton Keynes,
1982) remains a useful handbook, although it fails to make clear the sheer difficulty
of implementing a satisfactory record linkage routine.
57 R.J. Morris, Class, sect and party: the making of the British middle class - Leeds,
1820-50 (Manchester, 1990). Meanwhile R.J. Morris, Men, women and property in
England, 1780-1870 (Cambridge, 2005) adds rather more to the social history of the
propertied middle class than to the history of their political behaviour.
58 Phillips, Electoral behavior in unreformed England.
59 J.A. Phillips, The Great Reform Bill in the boroughs: English electoral behaviour,
1818-41 (Oxford, 1992), and J.A. Phillips and C. Wetherell, ‘The Great Reform Act
of 1832 and the political modernization of England’, American Historical Review,
100 (1995), pp. 411-36.
60 O’Gorman, Voters, patrons, and parties.
2.1 POLL BOOKS 31
61 V.W. Turner, The ritual process: structure and anti-structure (Ithaca, NY, 1977), p.
167, discusses rituals of status reversal ‘in which … groups or categories of persons
who habitually occupy low status positions in the social structure are positively
enjoined to exercise ritual authority over their superiors; and they, in turn, must
accept with good will their ritual degradation’.
62 William Paley, The principles of moral and political philosophy (2 vols, Dublin,
1785), ii, p. 197.
63 Functionalist anthropological accounts of such rituals, such as M. Gluckman, Rituals of
rebellion in south-east Africa (Manchester, 1954), are increasingly disfavoured. They
have been replaced by more historically aware accounts of the process of political
change, such as the model of oscillating equilibrium developed by E.R. Leach, Political
systems of highland Burma: a study of Kachin social structure (1954).
64 F. O’Gorman, ‘Campaign rituals and ceremonies: the social meaning of elections in
England, 1780-1860’, Past and Present, 135 (1992), pp. 79-115.
65 For the anthropological fascination with moments that mark changes of status, see
esp. A. van Gennep, Les rites de passage (Paris, 1908); translated by M.B. Vizedom
and G.L. Caffee as The rites of passage (1960).
66 M. Taylor, ‘Interests, parties and the state: the urban electorate in England, c. 1820-
72’, in Lawrence and Taylor (eds), Party, state and society, pp. 50-78, esp. pp. 52-3,
54.
67 Ibid., pp. 67-9.
68 On John Moody, see J.A. Hone, For the cause of truth: radicalism in London, 1796-
1821 (Oxford, 1982), pp. 63-4.
69 The implications of the Moody case are also discussed in section 1.10.2.
70 Corfield, Green and Harvey, ‘Westminster man’, pp. 157-85.
71 See details in section 6; and, for problems/challenges of classification, section 7.
72 Guildhall, Westminster poll book, 1784.
73 LMA Westminster Poll Books (1780), St Margaret and St John, WR/PP/1780/1-3; St
Anne, WR/PP/1780/4-5. These are the only parishes for which complete data survive
among the poll clerks’ books.
32 LONDON ELECTORAL HISTORY
74 LMA Westminster poll books (1788), WR/PP/1788/23.
75 WAC Westminster poll books (1796), E/3081A.
76 Guildhall, Westminster poll book, 1796.
77 Anon., A list of the persons who have polled for Humphry Parsons, Esq., Francis
Child, Esq., Aldermen, Richard Lockwood, Esq., Commoner, or one or more of them,
at the late election, for Members of Parliament, for the City of London [London,
1722] and ‘A list of the persons names who have polled for Robert Heysham, Esq.,
Alderman, Peter Godfrey, and John Barnard, Esqrs, at the late election of Members
of Parliament for the City of London, 1722’, Supplement to the Freeholder’s
Journal, 27 April 1722.
78 Anon., A poll of the livery-men of the City of London, at the election of members of
parliament, begun on Tuesday, April the 10th, 1722 (London, T. Payne, 1722).
79 This exercise replicates, for the sake of LED completeness, work done by W.A.
Speck and W.A. Gray, as reported in ‘London poll books, 1713’, in London Politics,
1713-17 (London Record Society, 17, 1981).
80 A list of disqualified voters may be found in Daily Courant, 3 Dec. 1713.
81 Daily Journal, 20 March 1724, ‘A list of the members of the several companies that
polled for Sir John Williams, Alderman, to be sheriff for the City of London, and
County of Middlesex’: BL Burney 242b. Anon., A list of the persons who have
polled for Sir John Williams, alderman, at the late election for sheriff for the City of
London and county of Middlesex (London, 1724): BL Hargrave Ms. 139, fos 167-82.
ESTC T 105993.
82 TNA SP/46/144, fos 30-32.
83 Daily Post, 7 Dec. 1724.
84 Anon., The poll of the livery of London for four citizens to represent the said City in
parliament (London, John Rivington, 1768). One of the copies at Guildhall has the
additional addenda leaf. This may date from 1770, when the death of William
Beckford triggered a parliamentary by-election and a reissue of the poll book.
85 Anon., A corrected list of the persons who have polled for the Rt. Hon. Frederick
Bull (London, Charles Rivington, 1773).
2.1 POLL BOOKS 33
86 Anon., A list of the persons who have polled for Richard Atkinson, Esq., which the
liverymen of the City of London are desired speedily and carefully to examine
[London, 1784]. ESTC T 200237.
87 Anon., A list of the liverymen of London, who voted for Mr Alderman Sawbridge,
and Richard Atkinson, Esq. at the late election for members of parliament for the
City of London, carefully corrected from the sheriff’s attested copies of the poll
(London, W. Lane, [1784]). ESTC T 198099.
88 See above, n. 45.
89 Anon., List of the streets and places within the administrative county of London
(1901), p. 355. This list indicates what ‘old name’ streets are contained within each
‘new name’, but unfortunately there is no index of ‘old name’ streets to identify, for
example, Nassau Street in Soho as the old name for Gerrard Place.
90 In the context of relatively limited opportunities for secure investment in the
nineteenth century, ‘building to let’ was a common phenomenon, and remained so
until the Rent Acts and Leasehold Reform Acts of the twentieth century intruded (for
good or ill) state regulation into the relationship between landlords and tenants.
91 In Rathbone Place, off Oxford Street, there survives today an ancient street sign
giving its name as ‘Rathbone’s Place’; another, in Gerrard Place, gives its name as
‘Wheeten’s Buildings in Nassau Street’.
92 D. Cruikshank, The secret history of Georgian London: how the wages of sin shaped
the capital (2009).
93 James Smith, Blacksmith, polled in 1727 from an address in Pissing Alley. Seven
years later James Smith, Blacksmith, polled giving his address as Bread Street, Star
Court, a cul de sac on the eastern side of Bread Street (and out of sight of the main
thoroughfare) which might have been Pissing Alley in its formal guise. However, A.
Harben, A dictionary of London (1918) suggests that the name was an old one for
Little Friday Street.
94 The numbering of houses within a street was given statutory force by 8 George III, c.
21 (1767), although this applied only to the City of London. For context, see P.J.
Corfield, ‘Walking the city streets: the urban odyssey in eighteenth-century Eng-
land’, Journal of Urban History, 16 (1990), pp. 149-50.
95 According to H. Phillips, Mid-Georgian London (1964), p. 6, rate collectors first
began to identify houses within each street by numbers in the period 1830-50, but
those in Westminster had preceded that. It should be noted that the apparently
34 LONDON ELECTORAL HISTORY
systematic numbering of houses shown in Richard Horwood’s Plan of London
(1799) bears little relationship with the systems of numbering used by rate collectors
or in the poll books.
96 Anon., List of the streets and places … (revised edn, 1912), p. v.
97 Anon., Copy of the poll ... for the city and liberty of Westminster (1780). The
publisher’s advertisement is on the verso of the title-page. Thomas Cornwall testified in
1789 that forty years before ‘they made a misentry with regard to my occupation ...
they called me carpenter instead of apothecary’: TNA PRO 30/8/237, fo 910.
98 The long-term decline in turnouts at British parliamentary general elections in the
second half of the twentieth century has belatedly focused attention on differential or
party-specific turnout as an important explanatory factor in political representation.
See H.D. Clarke, D. Sanders, M.C. Stewart, and P. Whiteley, Political choice in
Britain (Oxford, 2004), especially chapters 7 and 8.
99 [An alphabetical list of the livery-men of London, 1733], BL 1303 d.12.
100 Anon., A list of the liverymen of the several companies of the City of London
[London, 1750]. Although undated, this document records that Stephen Theodore
Janssen was sheriff, thus dating it to 1750. Inconsistencies between companies in the
recording of forenames and occupations of liverymen, and in the ordering of the
liverymen alphabetically or by seniority, suggest that it was compiled from lists
submitted by the companies. This list was used by G. Rudé in Wilkes and Liberty,
pp. 212-3, where it was dated to 1756.
101 Anon., A list of the names of the liverymen of the City of London (London, 1751).
102 Anon., A list of the livery of London alphabetically arranged (London, H.L. Galabin,
1792); and Anon., List of the whole body of the liverymen of London (London, J.
Wilkes, 1796).
103 [Anon., A list of the livery of London, alphabetically arranged under their several
wards, districts, and other places of residence (London [1802])]. This volume, in
IHR, lacks a title page. No other copy can be traced in NSTC.
104 Anon., A list of the liverymen of London (London, J. Miller, 1776), stated on the title
page that it was ‘collected from the last poll for chamberlain’. Meanwhile Anon., List
of the whole body of the liverymen of London (London, J. Wilkes, 1792) claimed to
have been ‘checked by the poll books of the last contested election’. Finally, Anon.,
A list of the livery of London alphabetically arranged (London, H.L. Galabin, 1796)
stated that it was ‘taken from the poll books of the last election’.
2.1 POLL BOOKS 35
105 Turnouts in London elections were slightly lower than the proportions estimated for
other places by O’Gorman, Voters, patrons, and parties, pp. 182-91, and by Phillips,
Great Reform Bill, pp. 32-6, probably because London’s franchise did not require
residence and distant electors had less incentive to vote than did local residents.
106 The City Elections Act (1725) made provision ‘as has been usual’ for requiring the
livery companies’ clerks to deliver lists of liverymen in the event of a controverted
election. The scrutiny following the by-election for sheriff of London and Middlesex
in 1724 began by calling for lists of liverymen from their respective company. These
were then checked for instances of multiple voting, of voting by those who were
disqualified by having had their livery fines returned, and of personation. See BL
Hargrave Ms. 139, fos 285-7.
107 See A catalogue of the several members of the Society of Apothecaries issued at
various dates throughout the first half of the eighteenth century.
108 Anon., A list of the court of assistants and livery of the worshipful company of
Drapers (annually from 1769); Anon., The names of the Master, wardens, assistants
and livery of the worshipful company of Tylers and Bricklayers (annually from
1772); Anon., A list of the wardens, assistants, and livery of the worshipful company
of Goldsmiths (annually from 1802); Anon, The names of the master, wardens,
assistants, and the rest of the livery of the worshipful company of clockmakers
(1802).
109 C.R. Webb (ed.), London livery company apprenticeship registers (41 vols, 1996-
2004).
110 28 George III, c. 36 (1788).
111 It was suspended by 29 George III c. 13 (1789), and repealed by 29 George III c. 18
(1789). See E. and A. Porritt, The unreformed House of Commons, i, pp. 26-8; and
O’Gorman, Voters, patrons, and parties, p. 132.
112 J. Prest, Politics in the age of Cobden (1977), pp. 1-9, contains a useful summary of
attempts to introduce electoral registration.
113 Land Tax duplicates for Middlesex may be found at LMA MR/PLT and WR/PLT.
Unfortunately virtually nothing survives for the period before 1780, and the registers
for 1767, used by George Rudé in his pioneering study of the Wilkite movement
(above, n. 45), are now deemed too fragile to be consulted.
114 42 George III c. 116, s. 200 (1802).
36 LONDON ELECTORAL HISTORY
115 For that reason Arthur Morris, the high bailiff and returning officer, noted that it was
difficult to establish precisely the total size of the electorate: see his submission to
Return of electors registered as qualified to vote in the last General Election in
Great Britain, BPP (1836), XLIII, p. 404.
116 Cited in W.A. Armstrong, ‘The use of information about occupation’ in E.A. Wrigley
(ed.), Nineteenth-century society: essays in the use of quantitative methods for the study
of social data (Cambridge, 1972), p. 210. See also discussion in R.J. Morris, ‘Occupa-
tional coding: principles and examples’, Historical Social Research, 15 (1990), pp. 3-
29.
117 This identification was revealed by linking records by name and address to poll
books from previous and successive elections.
118 ‘Comic and curious occupational titles from the 1881 census’, by Matthew Woollard
(2000) – copy from author.
119 For the problems and potentialities of studying occupations, see section 7.11.
120 For the Westminster scrutiny of 1749, see BL Lansdowne Ms. 509; WAC E/3079;
and Anon., Considerations on the determination of the scrutiny for the City and
Liberty of Westminster (London, J. Barnes, [1751]). This pamphlet is at ULL.
121 London Chronicle, 27 Oct. 1772.
122 Figures for the final declaration of the poll are in Morning Chronicle and London
Advertiser, 5 May 1784.
123 For the Westminster scrutiny of 1784, see BL Add. Ms. 36,226; Bodleian Ms. Eng.
Hist. c. 236; Anon., History of the Westminster election (2nd edn, 1785), pp. 539-74;
and Anon., Thoughts on the merits of the Westminster scrutiny, and the probable
causes of its institution (1785). Contemporary debates on the scrutiny shown in
Anon., Fox and Pitt’s speeches in the House of Commons (London, J. Debrett, 1784)
have been replicated in historians’debates between the Pittite P. Kelly, ‘Pitt versus
Fox: the Westminster scrutiny of 1784-5’, Studies in Burke and his time, 14 (1972-
3), pp. 155-62; and the Foxite L. Werkmeister, ‘Pitt versus Fox: a response to Paul
Kelly’, ibid, 15 (1973-4), pp. 45-50.
124 An act of 1770, known as Grenville’s Act after its sponsor (10 George III c. 16) had
attempted to regulate the composition of the parliamentary committee to remove
partisanship; but the issue remained under intermittent dispute until the Election
Petitions Act (1868) substituted the jurisdiction of a committee of judges.
2.1 POLL BOOKS 37
125 Each candidate in county elections was allowed an inspector by 7 & 8 William III, c.
25, s. 3 (1696). For cities and towns being counties in themselves, the right of the
inspector for each candidate to keep a check book of the poll was allowed by 19
George II, c. 28, s. 5 (1746), implying that the existence of such inspectors was by
that date commonplace.
126 BL Add. Ms. 33,123, fo 119.
127 Cited in L.G. Mitchell, Charles James Fox (Oxford, 1992), p. 70.
128 Cited in L. Reid, Charles James Fox: a man for the people (1969), p. 208.
129 Lists of those who ‘appear to have voted for’ Trentham in the Westminster by-
election of 1749 were produced by Vandeput’s committee to gain evidence for a
scrutiny.
130 CJ, 44, p. 125. An election print (George, no. 7,363) entitled ‘Six voters made out of
one, in favour of Lord John Townshend’ showed one man in six different disguises.
George also cites an advertisement pasted to the print in BM Department of Prints and
Drawings, offering for sale ‘all the curious and valuable wardrobe, in which all the
various voters masqueraded at the last election for Lord John Townshend’.
131 R. Grenville, Memoirs of the court and cabinets of George III (1853-5), i, p. 418.
132 TNA PRO 30/8/237, fo 785.